Streetscape skeleton measurement and classification

摘要

The scale and proportions of "streetscape skeletons," the three-dimensional spaces of streets defined by the massing and arrangement of surrounding buildings, are theoretically relevant to the way human users perceive and behave. Nonetheless, the dominant ways of measuring and identifying streets emphasize vehicular service and functionality. Moreover, existing built environment-based classifications have focused on recommended forms rather than characterizing the full range of existing conditions that must be accounted for in policy and understanding of human-environment interactions. To work toward a better streetscape measurement and classification scheme, this study investigated how large numbers of streetscapes could be efficiently measured to evaluate design patterns across and between multiple cities. Using a novel GIS-based method, 12 streetscape skeleton variables were measured on more than 120,000 block-length streetscapes in three northeastern U.S. cities: Boston, MA, New York, NY, and Baltimore, MD. Logistic regression models based on these variables were unsuccessful at distinguishing between cities, confirming that the variables were similarly applicable to each city and that the cities had comparable streetscape skeleton identities. Cluster analyses were then used to identify four streetscape skeleton classes that were also consistent between cities: upright, compact, porous, and open. These classes were distinct from the widely used highway functional class system, reinforcing the distinction between streetscape design and roadway functionality and thus the importance of accounting for them separately. The streetscape skeleton classes provide a digestible yet objective system for identifying prevalent streetscape designs that are appropriate for urban policy design, advocacy, and urban systems research.